|
(11) | EP 3 037 861 A1 |
| (12) | EUROPEAN PATENT APPLICATION |
|
|
|
|
||||||||||||||||||||
| (54) | Imaging method, and system, for obtaining an super-resolution image of an object |
| (57) Method for obtaining an super-resolution image (22) of an object (5), based upon
an optical microscope (21) including a support plate (6) for bearing the object, an
illumination source (1) for focusing an illumination beam (14) onto a target region
of the support plate, a digital camera (9) including a matrix of sensors, comprising
: - capturing, by the digital camera, a first image of the target region; - extracting, from the first image, a first block of pixel values provided by a sub-matrix (B0) of the matrix of sensors; - displacing, by a sub-diffraction limited distance, the support plate by the displacement block along a displacement axis; - capturing, by the digital camera, a second image of the target region; - extracting, from the second image, a second block of pixel values provided by the sub-matrix (B0) of the matrix of sensors; - storing said first and second blocks of pixel values as a first and second blocks of pixel values to be placed right next each other in the super-resolution image along the image axis (X, Y) corresponding to the displacement axis. |
[0001] The present invention generally relates to the field of the production of magnified images of small objects using optical microscopes.
[0002] More specifically the invention relates to an imaging method for obtaining an super-resolution image of an object, based upon an optical microscope adapted for capturing a image of the object, said optical microscope including :
a support plate for bearing the object,
an illumination source adapted to produce an illumination beam for illuminating the object,
an optical element for focusing the illumination beam on the support plate, the section of the object on the support plane currently illuminated by the focused illumination beam being called hereafter the target region,
a digital camera including a matrix of sensors for capturing an image of the target region, each sensor providing for a respective pixel value and
a displacement block for displacing the support plate relative to the focused illumination beam and to the digital camera along at least two displacement axes among three axes x, y, z perpendicular to each other, wherein the axes x and y define the plane of the support plate, said at least two displacement axes defining two corresponding perpendicular image axes of the super-resolution image.
[0003] The resolution of an optical microscope relates to the ability of the microscope to reveal adjacent details as distinct and separate. Independently of imperfections of lenses, the optical microscope's resolution is limited by light's diffraction.
[0004] Indeed, because of light's diffraction, the image of a point is not a point, but appears as a fuzzy disk surrounded by a diffraction, called "Airy disk" or "point spread function". Thus two points of the object, adjacent but distinct, will have for image points two spots whose overlapping can prevent from distinguishing the two image points : the details are then not resolved.
[0005] According to the Abbe diffraction limit, the limit of resolving separate points by the optical microscope, known as the diffraction limit, is stated as
where λ is the wavelength of the illumination source, and NA is the numerical aperture of the objective lens of the optical microscope. In general, the lowest value of the diffraction limit obtainable with conventional objective lens is 200 nanometers (nm).
[0006] There are known processes for producing images with higher resolution than the resolution allowed by simple use of diffraction-limited optics : for example, the Stimulated Emission Depletion microscopy (STED), the Spatially Structured Illumination Microscopy (SSIM), Photoactivated Localization microscopy (PALM), the Stochastic Optical Reconstruction Microscopy (STORM).
[0007] Documents relative to such techniques are for example WO 2013/153294 A1, or "Fluorescence microscopy with diffraction resolution barrier broken by stimulated emission", Klar et al., Proc Natl Acad Sci USA. 2000 Jul 18;97(15):8206-10.
[0008] A common thread in these techniques is that they are able to resolve details beyond the diffraction limit, for example in PALM this is achieved by switching fluorophores on and off sequentially in time, so that the signals can be recorded consecutively.
[0009] Unfortunately, these super-resolution methods require purchasing expensive optical platforms (e.g. more than 1 M€ for STED) and/or require further significant post signal data treatment, both being beyond the resources of most cellular biology laboratories.
[0011] According to a first aspect, the invention proposes an imaging method for obtaining a super-resolution image of an object as mentioned here above, said method being characterized in iterating the following set of steps :
- capturing, by the digital camera, a first image of the target region currently illuminated by the focused illumination beam;
- extracting, from the captured first image, a first block of pixel values provided by a sub-matrix of the matrix of sensors;
- storing said first bloc of pixel values as a first block of pixel values of the super-resolution image;
- displacing, by a sub-diffraction limited distance, the support plate by the displacement block along one of the displacement axis;
- capturing, by the digital camera, a second image of the target region currently illuminated by the focused illumination beam;
- extracting, from the captured second image, a second block of pixel values provided by said sub-matrix of the matrix of sensors;
- storing said second bloc of pixel values as a second block of pixel values of the super-resolution image, said second block of pixel values being placed right next to said first block of pixel values in the super-resolution image along the image axis corresponding to said one displacement axis.
[0012] Our analogous method to the aforementioned techniques provides for a simple way to get a super-resolution image of an object with limited calculations and treatments.
[0013] According to some embodiments of the invention, an imaging method for obtaining a super-resolution fluorescence image of an object includes further the following features:
- the sub-matrix comprises only one sensor;
- a sensor of the matrix is selected as sensor of the sub-matrix in a preliminary step
only if said sensor captures a part of the image of the focused illumination beam
on the support plate;
- the sensor(s) of the matrix is/are selected as sensor(s) of the sub-matrix in a preliminary step as the one(s) among the sensors of the matrix capturing a part of the image of the focused illumination beam on the support plate wherein the image is the most in focus;
- the method comprises a step of applying a background removal to the pixel values of the super-resolution image and/or including a step of cropping said pixel values that are outside at least one specified range of upper intensity values, in order to remove from the pixel values related to an individual fluorophore the overlapping intensities of neighbouring fluorophores;
- the object includes at least one fluorescence source, and the illumination source is adapted to produce an illumination beam for generating fluorescence from the illuminated object, the image of the target region captured by the digital camera being an image of the fluorescence generated by the illuminated object, the obtained super-resolution image being a super-resolution fluorescence image;
[0014] According to a second aspect, the invention proposes a system for obtaining a super-resolution image of an object, said imaging system including a controller and an optical microscope, said optical microscope including :
- a support plate for bearing an object to be imaged,
- an illumination source adapted to produce an illumination beam for illuminating the object,
- an optical element for focusing the illumination beam on the support plate, the section of the object on the support plane currently illuminated by the focused illumination beam being called hereafter the target region,
- a digital camera including a matrix of sensors for capturing an image of the target region, each sensor providing for a respective pixel value, and
- a displacement block for displacing the support plate relative to the focused illumination beam and to the digital camera along at least two displacement axes among three axes x, y, z perpendicular to each other, wherein the axes x and y define the plane of the support plate, said at least two displacement axes defining two corresponding perpendicular image axes of the super-resolution image;
[0015] The present invention is illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
- Figure 1 shows a schematic lateral view of a system for obtaining a super-resolution fluorescence image of an object in an embodiment of the invention;
- Figure 2 is a representation of a top view of the platform supporting the sample;
- Figure 3 is a representation of a bottom view of the camera sensor matrix;
- Figure 4 is a view of a super-resolution image obtained according to an embodiment of the invention;
- Figure 5 represents steps of a method for obtaining a super-resolution fluorescence image of an object in an embodiment of the invention;
- Figure 6 is a schematic view of an ordinary fluorescence scanning microscope;
- Figure 7 is a schematic view of an ordinary confocal fluorescence scanning laser microscope;
- Figure 8 is a schematic view of a fluorescence scanning microscope according to the invention;
- Figure 9 shows the cross sections of the normalized intensity in the detector plane for two single point fluorescence sources.
Ordinary fluorescence scanning microscopy
[0017] In Ordinary fluorescence scanning microscopy (OFSM), the laser light is focused into a diffraction limited spot exciting the fluorescence of the objects that are inside that spot. The light emitted as the fluorescence is collected by the photo-detector. Then, the laser beam scans the surface of the sample; usually by moving the sample with x-y-(z) translation stage relative to the laser beam. The final fluorescence image is reconstructed from step-by-step measured fluorescence intensities. A generalized scheme of OFSM technical set-up is shown in Figure 6, showing a fluorescence scanning microscope including a sample plane 31, a diffraction limited laser spot 32, an optical element 33 such as filters cubes and/or mirrors, a detector plane 34, a diffraction limited laser spot's image 35 on the detector plane.
[0018] Assuming that there are two single point (infinitely small) sources of fluorescence light on the sample plane and that the wavelength of light is 500 nm and that numerical aperture (NA) of our system is 1.45, therefore the lateral resolution of such system in ideal conditions is approximately equal to 176 nm (0.51*500nm/1.45), based on the Abbe diffraction limit. Note, that as a general rule the light emitted by two independent fluorescence sources will not be coherent. The image produced by these two objects on the detector plane will represent two circular light spots with surrounding concentric circles with much lower intensity (called Airy patterns). Thus each of these fluorescence sources will produce a typical diffraction image of the single point source on the detector plane comprising a central bright part and surrounding concentric rings.
[0019] The main intensity of the fluorescence will be accumulated inside a central bright part, which in general can be nominally described as a Gaussian distribution. In the following description, the influence of the concentric rings will be neglected and the only central part will be taken into account. Moreover, in case of OFSM small size detectors as often used, so mainly central part of the image is only collected for step-by-step fluorescence image reconstruction.
[0020] The collected signal for a single scanning step in case of OFSM will be the integrated fluorescence intensity of the total fluorescence signal hitting the detector while measuring that point. In the ideal case it is only the centre of the diffraction image. In the considered assumption (wavelength of light of 500 nm, NA=1.45) and under ideal experimental conditions the cross section of the central part of the signal collected can be represented as a Gaussian with the full width at half maximum (fwhm) of 176 nm. Consequently, the total signal intensity for the single point in the future fluorescence image is the signal limited by fwhw=176 nm Gaussian distribution. In order to determine the resolution of the system in these ideal conditions, behaviour of the two Gaussian distributions can be checked in order to determine when they can no longer be distinguished as individual objects.
[0021] In order to achieve this, one can examine the second derivative of the overlapping Gaussians (which is a more precise method). In a more simplistic approach it can be looked at how the regions of the fluorescence signal registrations overlap. It is intuitive that a normal OFSM microscope will never reach the above theoretical resolution of 176 nm (unless the detector is small and rightly positioned). In addition, it can be seen from the cross sections of two single point sources on the detector plane for the OFSM registration conditions that there is no simple way to resolve two single point sources in the ideal case for OFSM even at the 200 nm distance.
Confocal fluorescence scanning laser microscopy
[0022] The main difference of confocal fluorescence scanning laser microscopy to that of the aforementioned OFSM is the insertion of an additional small size pinhole which removes/blocks part of the spectral contamination due to the presence of the additional rings from the Airy disk, thus the detector receives only the central part of the fluorescence that was excited by the incident laser beam. As a general rule, the best resolution is obtained in case when one uses a pinhole equal to one Airy unit.
[0023] A schematic confocal fluorescence scanning laser microscope is shown in figure 7 including an additional pinhole 36 compared to the microscope of figure 6 and thereby providing a diffraction limited laser spot's image 39 on the detector plane 34. As shown, the diffraction limited laser spot's image 39 is only a section of the diffraction limited laser spot's image 35 obtained with the microscope of figure 6.
[0024] The aforementioned example of two single point fluorescence sources as observed by an OFSM with the insertion of a one Airy unit pinhole in the optical path (the bigger the pinhole the lower is the resulting image quality and resolution) is now re-examined in regard to such a OFSM. The registered intensity of the signal will be mainly the integral intensity area limited between bars, where distance between the bars equals the fwhm. Therefore, in contrast with OFSM two single point fluorescence sources lying 200 nm apart should be easily resolvable in a confocal regime with a 1 Airy unit pinhole. Thus, the theoretical predicted resolution can be achieved.
Super-resolution microscopy according to the invention
[0025] The present invention goes some steps further:
- i) For the image reconstruction a single pixel (or a single block of pixels) is used that is located within the central area of the diffraction limited spot on a detector plane and which is smaller than the diffraction limited spot on the detector plane.
- ii) A (flat) background is removed in an embodiment. Consequently, the bulk of the fluorescence signal, which originates from the object lying in the geometrical optical axis of our pixel, is used for image reconstruction.
[0026] Employing these two steps a significantly higher resolution is achieved, and avoid the use of complicated data acquisition methods and subsequent analytical procedures. It is obvious that not only one pixel is available for image reconstruction, a few pixels can be used successfully, producing shifted images available for averaging. Moreover, the possibility for three dimensional picture recovery from the single scan exists because every pixel probes at a different geometric point within the three dimensional object under investigation. The generalized scheme of the technical set-up is shown in figure 8. showing a schematic microscope according to the invention showing a single pixel 37 inside the diffraction limited laser spot's image 35 in the detector plane 34 used for reconstructing the final image, and showing also a geometrical projection 38 of the detector's pixel 37 on the sample plane 31.
[0027] In addition, scanning with significantly precise and small steps is a requisite. However step size can be relatively large in cases where we use images obtained from multiple pixels for averaging (with necessarily corrections), therefore making the overall scanning faster.
[0028] For example, looking into the aforementioned case in an example of a fluorescence microscopy, and referring to Figure 9 showing the cross sections of the normalized intensity in the detector plane for two single points fluorescence sources, it is possible to resolve two single point sources of fluorescence lying at a distance of 100 nm by applying a "detector" pixel size of ∼50 nm (here assumed to be circular) and 30% flat background removal. Greater background removal is often better, but it is always a trade in with signal to noise ratio. Smaller pixel sizes lead to lower background removal sizes and better resolution and signal to noise ratio.
[0029] Thus the present invention is analogous to other super-resolution imaging techniques but not based on them.
[0030] This present invention is able to obtain super-resolution images under high magnification that is comparable with other known techniques.
[0031] This present invention is able to increase the resolution of any image, irrespective of the value of magnification of the sample. Hence, this present invention is able to improve images at low-magnification hitherto considered to be not suitable for high-resolution imaging.
[0033] A system 20 for obtaining a super-resolution fluorescence image of an object in an embodiment of the invention is schematically represented in Figure 1.
[0035] The optical microscope 21 includes an illumination source 1, a spatial filter 2, a dichroic mirror 3, an objective 4, a platform 6, an emission filter 7, a combination 8 of lenses, for example achromatic, and a digital camera 9.
[0037] The optical microscope 21 can be a microscope usually used for capturing fluorescence images, such as the Nikon Ti Eclipse ® for example.
[0039] The sample 5 to be imaged is adhered to a standard coverslip, or other suitable support, that is fixed rigidly by a sample holder on the platform 6.
[0040] The sample's surface on the platform 6 lies on a plane define by the perpendicular axes x, y.
[0042] According to classical fluorescence imaging, tiny fluorescence sources, for example fluorophores based upon the one photon fluorescence physical phenomenon, has been attached to the sample 5. These fluorescence sources can be intrinsic to the sample.
[0043] The illumination source wavelength is chosen to induce fluorescence from the attached fluorescence sources: the fluorescence sources, when lightened by the illumination source, absorb the energy of the illumination source's wavelength(s) and emits fluorescence light, whose wavelength is higher than the illumination source's wave length(s).
[0044] The controller 10 is adapted for defining displacement commands and providing the displacement commands to the motion block 11.
[0045] The displacement commands are defined by the controller 10 as a function of a predetermined Region of Interest (ROI) of the sample 5. A displacement command commands the motion block 11 to displace the platform 6 and specified the commanded displacement along the axes x, y, z, z being an axis perpendicular to the axes x, y.
[0046] The controller 10 is further adapted to trigger the capture of an image by the digital camera 9 by sending a capture command to the digital camera 9, and the controller 10 is adapted to construct the super-resolution image of the sample ROI from successive fluorescence images captured by the digital camera 9, as detailed hereafter.
[0047] In an embodiment, the controller 10 includes a microprocessor and a memory (not represented). The memory stores software instructions 10, which, when executed by the microprocessor, produce the operation of the controller 10.
[0048] The digital camera 9, for example includes a two-dimension (2D) matrix of sensors wherein each sensor is adapted to detect a local fluorescence intensity during a determined time.
[0049] The digital camera 9 is adapted to capture an image upon reception of a capture command from the controller 10.
[0050] When the digital camera 9 is capturing an image of a sample currently illuminated by the illumination source 1, each sensor detects a respective local fluorescence intensity and the camera 9 then determines, for each pixel corresponding to each sensor, a respective pixel intensity as a function of the local fluorescence intensity detected by the sensor.
[0051] The digital camera 9 is then adapted for providing an image composed of a matrix of pixels (the position of a pixel in the matrix of pixels is the same as the position of the corresponding sensor in the matrix of sensors).
[0052] Figure 3 represents a bottom view of the sensor matrix of the camera 9. Said sensor matrix has L ranks of sensors parallel to the axis x and I columns of sensors parallel to the axis y. This view of Figure 3 is also considered as representing the matrix of the L x I pixels of an image captured by the camera 9, wherein I is the size of the matrix along an image dimension X corresponding to the axis x, and L is the size of the matrix along an image dimension Y corresponding to the axis y.
[0053] The motion block 11 is adapted for receiving displacement commands from the controller 10 and for, upon reception of such a command defining a commanded displacement, moving the platform 6 by the commanded displacement
[0056] If necessary, as a function of wavelength involving fluorescence from the sample (excitation wavelength), the emitted light is cleaned by the optional LASER filter 1a in order to reject, from the emitted light, the wavelengths distinct from the excitation wavelength.
[0058] The exciting light beam of thus collimated incident light passes through the dichroic mirror 3 (as known, the dichroic mirror 3 lets exciting beam to fall on the platform 6 and reflects fluorescence light to the digital camera 9) and through the objective 4 in order to produce a focused diffraction-limited spot 14 on the sample 5 supported by the platform 6.
[0059] The fluorescence light, emitted by the area of the sample 5 covered by the spot 14, returns along the incident optical path, via the objective 4, which magnifies the sample image, to the dichroic mirror 3, and continues onwards to the emission filter 7 that keeps only the fluorescence wavelength(s) of interest.
[0060] The combination 8 (of optional) lenses further magnify the fluorescence light provided onto the camera sensor matrix 9.
[0061] The sensor matrix of the camera 9 is positioned in regard to the spot 14 so that the fluorescence light emitted by the sample area in the spot 14 is collected towards the sensor matrix.
[0062] As known, fluorescence measurement requires that a pulse of incident light arrives at the sample. This is obtained by placing a shutter (not shown) at the exit of the source 1.
[0064] Such focus is preliminary obtained for example by a piezo device attached to the objective 4, or by commanding a displacement along the axis z of the platform 6 by the controller 10 to the motion block 10.
[0065] In an embodiment, a LASER (Light Amplification by Stimulated Emission of Radiation) is used as the illumination source 1, either of constant intensity (Continuous Wave - CW) or pulsed.
[0066] In an embodiment, the magnification of the objective 4 for the fluorescence light is in the range that is defined by the specifications of the digital camera and the zoom of the combination 8 of lenses as required.
[0067] The digital camera 9 is for example a 2D detector for example an (Electron Multiplying) Charge Coupled Device ((EM)CCD) type, or a CMOS (Complementary Metal Oxide Semiconductor). The 2D detector may be replaced by a 1 D detector, or point detector, if required. In this latter case, B0 is equal to I x L, in the description described below.
[0069] In an embodiment, each EMCCD pixel is a square of s x s µm2 where s is in the range [6 µm, 15 µm].
[0070] For example, when using an objective of NA=1.45, the diameter of the spot 14 in the plane (x,y) is in the range [150 nm ,350 nm].
[0071] These here above values are of course given only as examples and are dependent on the type of camera used.
[0072] In an embodiment, a combination of illumination sources may be used such that the collimated beams all have the same trajectory when arriving at the dichroic mirror 3. In order to remove any potential artefacts due to different polarities of the laser beams they are depolarised incident to the dichroic mirror 3. Firstly, the polarity of all the laser beams is orientated to the same direction by the linear polarizer 1 b. Secondly, an assembly 1 c of a circular polarizer coupled to an achromatic depolarizer 1c is inserted into the optical path immediately after the spatial filter 2. Alternatively, if the combination of wavelengths and laser output powers are too diverse, each laser beam will have its own spatial filter 2 prior to arrival at the depolarizer 1 c. The beam will thus converge after the spatial filters 2. If the use of fully depolarised beams is not considered to be required this option may be removed from the optical path, leaving only the spatial filter(s) 2.
[0073] According to an embodiment of the invention, the following steps are implemented based on the system 21.
[0074] In the considered embodiment, the spot 14 does not move relative to the camera 9 during the construction of a super-resolution image according to the invention.
[0075] In a first step 101, a block of sensors B0 is chosen among the I x L sensors of the camera 9. The number of sensors n inside B0 is strictly less than I x L.
[0077] The sensors of the block B0 form the sensor matrix which may be any shape (square, rectangle, oval, linear array, single point, etc.). Hereon, we shall consider the case of the block B0 being a square of dimension n.
[0078] For example, n is equal to 1, or any other size that is deemed suitable. The value of n is for example less than 4, 16, 25.
[0079] The block of sensors B0 is further chosen so that the block of pixels corresponding to the block B0 of sensors is located inside the image, referenced A in Figure 3, of the spot 14 in any captured image. It means that the size of the block of pixels corresponding to the block of sensors B0 is smaller than the size of the diffraction spot 14's image (the block of pixels corresponding to the block of sensors B0 is indeed inside the diffraction spot 14's image).
[0080] In the following, n=1 is considered: B0 comprises only the (n0, m0)sensor of the matrix of sensors, i.e. the sensor of position (n0,m0) in the matrix of sensors.
[0081] In a step 102, a ROI of the sample 5 is defined, for example the ROI defined by the rectangle O1, O2, O3, O4 in the plane of the Figure 2's top view, and the definition of this ROI is provided to the controller 10.
[0082] For example, the ROI bounds an entire prokaryote, or small subcellular structure, cf. actin filaments, in a larger biological sample 5.
[0083] A value of distance step is defined, called d hereafter. d is for example chosen by an operator.
[0084] The value of d is strictly smaller than the diffraction limit of the optical microscope 21 :
where λ is the wavelength of the illumination source 1, and NA is the numerical aperture of the lens in the objective 4.
[0085] For example, in case the diffraction limit is 200 nm, d is chosen in the range [15 nm, 50 nm].
[0086] Then the controller 10 will commands the progressive scanning of the ROI by steps of distance d along the axis x, or along the axis y, from an initial point of the ROI in reiterated steps 103 and 104 as detailed hereafter.
[0087] At first, an initial displacement of the platform 6 is determined by the controller 10 in regard to the current position of the platform 6.
[0088] In the considered embodiment, the initial displacement of the platform 6 is determined by the controller 10 in order to firstly direct the focused diffraction-limited spot 14 on the point O1.
[0089] A displacement command is provided by the controller 10 to the motion bloc 11, indicating the determined initial displacement.
[0090] Upon reception of this displacement command, the platform 6 is positioned by the motion block such that the spot 14 covers the point O1.
[0091] In a first iteration of step 103, the controller 10 triggers the capture of an image by sending to the camera 9 an image command whereas the spot 14 covers the point O1.
[0092] Upon reception of this capture command, the camera 9 captures an image as a function of the currently fluorescent light collected by the sensor matrix.
[0093] The controller 10 extracts, from the captured image, the block of pixels corresponding to B0 at this first iteration. Thus in the considered embodiment, it extract the intensity of the pixel P(n0,m0) corresponding to the (n0,m0) sensor (i.e. the pixel of position (n0,m0) in the matrix of pixels).
[0095] Said intensity Int(n0,m0)1,1 is stored as the intensity of the pixel of position (1,1) in the matrix of pixels of the super-resolution image 22, referring to Figure 4.
[0096] Then in a first iteration of step 104, the controller 10 commands to the motion block 11 a displacement of d along to the axis x from the current position, so that the spot 14 be displaced relative to the sample 5 on platform 6, along x in direction of the limit of the ROI defined by [O2, O3].
[0097] And the motion block 11, upon reception of the displacement command, displaces the platform 6 accordingly to the received displacement command.
[0098] Then, in a second iteration of step 103, the controller 10 triggers the capture of an image by sending to the camera 9 an image command.
[0099] Upon reception of this capture command, the camera 9 captures an image as a function of the currently fluorescent light collected by the sensor matrix (emitted from an area of the sample shifted by distance d from O1 along the axis x.
[0100] The controller 10 extracts, from the captured image, the block of pixels corresponding to B0 at this second iteration of step 103, thus in the considered embodiment, it extract the intensity of the pixel P(n0,m0) corresponding to the (n0,m0)-sensor.
[0102] Said intensity Int(n0,m0)1,2 is stored as the intensity of the pixel of position (1,2) in the matrix of pixels of the super-resolution image 22, referring to Figure 4.
[0103] In the super-resolution image 22, this pixel is adjacent according, to the axis X, to the pixel of intensity Int(n0,m0)1,1 previously stored.
[0104] Then in a second iteration of step 104, the controller 10 commands to the motion block a displacement of d along to the axis x from the current position so that the spot 14 be displaced relative to the sample 5 on platform 6, along x in direction of the limit of the ROI defined by (O2, O3).
[0105] And the motion block 11, upon reception of the displacement command, displaces the platform 6 accordingly to the received displacement command.
[0106] The group of steps 103 and 104 is reiterated along the axis x until the Nth iteration, wherein the position of platform is such that the spot 14 directs to point O2.
[0107] At said Nth iteration of step 103, the intensity Int(n0,m0)1,N of the pixel P(n0,m0) is stored as the intensity of the pixel of position (1,N) in the matrix of pixels of the super-resolution image 22. In the super-resolution image 22, this pixel being adjacent according to the axis X to the pixel of intensity Int(n0,m0)1,N-1 previously stored.
[0108] The controller 10 detecting that that the limit of the ROI along the axis x has been reached, in the Nth iteration of step 104, the controller 10 commands to the motion block 11 a displacement command of d along to the axis y this time, from the current position so that the spot 14 be displaced from O2 relative to the sample 5 on platform 6, along y in direction of the limit of the ROI defined by (O3, O4).
[0109] And the motion block 11, upon reception of the displacement command, displaces the platform 6 accordingly to the received displacement command.
[0110] Then at the N+1th iteration of step 103, the intensity Int(n0,m0)2,N of the pixel P(n0,m0) is stored as the intensity of the pixel of position (2,N) in the matrix of pixels of the super-resolution image 22.
[0111] In the super-resolution image 22, this pixel is adjacent according to the axis Y to the pixel of intensity Int(n0,m0)1,N previously stored.
[0112] Then at a N+1th iteration of step 104, a displacement of d along to the axis x from the current position so that the spot 14 be displaced relative to the sample 5 on platform 6, along x in direction of the limit of the ROI defined by (O1, O4) is commanded by the controller 10 to the motion block 11.
[0113] And the motion block 11, upon reception of the displacement command, displaces the platform 6 accordingly to the received displacement command.
[0114] Successive iterations of the group of steps 103 and 104 are applied, displacing the spot 14 by step of d along the axis x towards the side (O1, O4) of the ROI, thus constructing the second rank of pixels of the super-resolution image 22, from Int(n0,m0)2,N to Int(n0,m0)2,1 reached at the 2Nth iteration of step 103.
[0115] Then at a 2Nth iteration of step 104, similarly to the explanations here below, he controller 10 then detects that the limit of the ROI along the axis x has been reached, and consequently in the 2Nth iteration of step 104, the controller 10 commands to the motion block 11 a displacement command of d along to the axis y from the current position so that the spot 14 be displaced from the current position along y in direction of the limit of the ROI defined by (O3, O4).
[0116] And the motion block 11, upon reception of the displacement command, displaces the platform 6 accordingly to the received displacement command.
[0117] And similar iterations are processed until all the ROI have been scanned by the spot 14, by successive displacements of distance d each one.
[0118] In the considered case wherein the ROI is rectangle, the resulting super-resolution image 22 comprises M ranks and N columns of pixels, wherein
et M =
wherein E is the function "integer part" and ∥PP'∥ is the function providing the distance between the two points P and P'.
[0119] But in other embodiments, the ROI can have various shapes and the controller 10 is adapted to scan the whole ROI and to construct the super-resolution image for this ROI.
[0120] In an embodiment, the scan in a plane (x,y) through sub diffraction displacements and the construction of the super-resolution image by successive extracted pixel block according to the invention is repeated for each plane z if collected (i.e. for successive layers through the sample, thus constructing a 3D super-resolution image).
[0121] Here below, the provided super-resolution image correspond to a scan in the plane defined by the axes x and y. In another embodiment, the scan and construction of super-resolution image is achieved in a plane defined by the axes x and z, or by the axes y, and z. Then the sub-diffraction displacements are achieved along the considered axes x and z or y and z, instead of the axes x and y as described here below.
[0122] In the detailed embodiment disclosed here below, the number n in the block B0 was considered to be equal to 1.
[0124] Thus when a first image and a second image are captured in successive steps 103, the position of the platform 6 for the first image being separated from the position of the platform 6 for the second image by a displacement d less than the diffraction limit in a direction along the axis x (respectively y), the super-resolution image will include a first block of pixels, said first block of pixels being the block of pixels corresponding, in the first image, to the block B0 of n sensors. And the super-resolution image further includes, adjacent to said first block of pixels in the direction along the axis X (respectively Y), a second block of pixels, said second block of pixels being the block of pixels corresponding, in the second image, to the block B0 of n sensors etc.: the super-resolution image is constructed by the blocks of pixels successively captured by the block B0 after each sub-diffraction displacement. After the series of images are collected and the fluorescence image is reconstructed as stated (steps 103-104), in some embodiments a further step of removing a (flat) background is implemented if required ; the data of the fluorescence image are then in an embodiment cropped such that only a specified range of upper fluorescence intensity values are plotted. This allows one to better discriminate between individual fluorophores by removing/reducing the overlap of the fluorescence signal from neighbouring emitters.
[0125] For clarity, we consider that the fluorescence image which has been reconstructed from the acquisition protocol (steps 103-104) has a normalized intensity range from 0 (minimum) to 1 (maximum). In the simplest case, a simple background can be removed ( for example, for each 2D sub-image of the super-resolution image, an estimation of a background noise intensity, estimated in an embodiment from an image obtained from a non fluorescence area) is subtracted to the pixel intensities of the 2D sub- image. Then the resulting fluorescence image can be scaled from 1 to X%, where X represents a user-defined normalized intensity value less than 1: the intensity value below X% are replaced by a value equal to zero, the intensity value V beyond X% are transformed into (1-X%)V.
[0126] This results in a plot that represents only the top X% (for example 30%) of the reconstructed signal, where the different fluorophores are deemed not to greatly interfere with each other, as in the example illustrated in Figure 9. The removal of the background, in particular the value of X, is specific to each sample of interest. Such a value is estimated in a preliminary stage.
[0127] Alternatively, a series of normalized intensity slices can be recreated, ranging from 1 to Xa, Xa to Xb, Xb to Xc, etc., where a,b,c, etc. represent user defined intensity values.
[0128] The sample considered in the embodiment detailed here below is a biological sample, but the sample could be non-biological in another embodiment.
[0129] According to the embodiment described here below, the sample to be imaged is moved relative to a fixed beam of collimated light and to a fixed camera. But in another embodiment, the sample is fixed whereas the motion is applied to the assembly of the collimated beam and the fixed camera.
[0131] In an embodiment, given the size n of the block B0, the sensor(s) of the block B0 in the step 101 are chosen according to the following way: the sample 5 is displaced so that the spot 14 is directed on a test area of the sample 5) containing fluorescence beads (or a different test sample containing fluorescence beads is used). An image is captured by the camera 9. The sections of image provided by all the blocks of n pixels inside the image of the spot 14 are analysed in order to determine the section that is the most in focus by examining the quality of the signal, and, for example, with reference to fluorescence properties of known samples. Often fluorescent microspheres (with defined diameters of between ca. 50 and 2000 nm) which provide a near perfect Gaussian distribution (when not aggregated) of fluorescence intensity are used to determine the section that is the most in focus. The block of sensors corresponding to the block of pixels with the best focus is then chosen as the block B0.
[0132] This disposition prevents from any additional scattered light that would blemish the super-resolution image.
[0133] The quality of the super-resolution image will be improved if the fluorescence sources of the sample 5 result in discrete sources of emission, and are not so concentrated that it results in a carpet of fluorescence intensity.
[0134] The method according to the invention was applied to the H-nS protein, a DNA-binding protein, as sample, which is tagged with the fluorophore GFP (Green Fluorescent Protein), in Escherichia (E.) coli cells. The images obtained according to the invention results present a superior resolution compared to images obtained using a super-resolution technique called STORM according to the results published in Wang et al., Science (2011) 333, 1445-1449).
[0135] The resolution of the resulting XY(Z) image is dependent on the increment d of the sub-diffraction-limited mechanical displacement, as well as the pixel size of the 2D detector. For example a resolution superior to 40 nm, as defined by Donnert et al., Proc Natl Acad Sci USA. 2006 Aug 1;103(31):11440-5, is obtainable with an excitation wavelength of 488 nm with the following hardware: optical microscope (21) - Nikon Ti; objective lens (4) - Nikon 100x CFI Plan Apochromat; displacement block (11) - Physik Instrumente; nanopositioning stage P-733 and digital camera (9) - Andor iXon Ultra 897.
[0136] Super-resolution according to this embodiment of the invention is basically achieved by conjugating a normal epi-fluorescence microscope (installed in a vibration-free environment) with a high-precision moving platform under the same experimental conditions as classical fluorescence microscopy, and with the proper software to reconstruct the images: it thus makes super-resolution accessible to a very large number of laboratories.
[0138] Of course, the invention is implementable for producing super-resolution images based on different types of microscopy: for example, transmission, scattering, or absorption microscopy.
[0139] In fluorescence microscopy the illumination light (excitation wavelength(s)) and captured light (emission wavelength(s)) are not the same, whereas in transmission, T, microscopy and absorption, A, microscopy they are identical. The ability to transmit the light is expressed in terms of an absorbance, A, which is normally written as
where I and 10 are defined as the intensity (power per unit area) of the illumination light (termed incident radiation) and the captured light (termed transmitted radiation), respectively. In scattering microscopy, the illumination light is reflected (elastically or inelastically) by the object and as captured light registered by the 2D detector.
[0140] Thus the invention enables to obtain a image of a sample with an effective resolution superior to that of the Abbe diffraction limit, by means of sub-diffraction length displacements in two (x, y), or three (x,y,z), dimensions while the object of study is illuminated with a focused light source. During data acquisition, the sample of study is housed on a xy(z)-stage which moves relative to a fixed beam of collimated light. Alternatively, the xy(z)-stage may be fixed in the x and y axes while the laser beam scans the object of study, by means of adaptive optics such as resonant scanning mirrors.
1. An imaging method for obtaining an super-resolution image (22) of an object (5), based
upon an optical microscope (21) adapted for capturing a image of the object, said
optical microscope including :
a support plate (6) for bearing the object,
an illumination source (1) adapted to produce an illumination beam for illuminating the object,
an optical element (4) for focusing the illumination beam on the support plate, the section of the object on the support plane currently illuminated by the focused illumination beam (14) being called hereafter the target region,
a digital camera (9) including a matrix of sensors for capturing an image of the target region, each sensor providing for a respective pixel value and
a displacement block (11) for displacing the support plate relative to the focused
illumination beam and to the digital camera along at least two displacement axes among
three axes x, y, z perpendicular to each other, wherein the axes x and y define the
plane of the support plate, said at least two displacement axes (x, y) defining two
corresponding perpendicular image axes (X, Y) of the super-resolution image (22);
said method being characterized in iterating the following set of steps :
- capturing, by the digital camera, a first image of the target region currently illuminated by the focused illumination beam (14);
- extracting, from the captured first image, a first block of pixel values provided by a sub-matrix (B0) of the matrix of sensors;
- storing said first bloc of pixel values as a first block of pixel values of the super-resolution image;
- displacing, by a sub-diffraction limited distance, the support plate by the displacement block along one of the displacement axis;
- capturing, by the digital camera, a second image of the target region currently illuminated by the focused illumination beam (14);
- extracting, from the captured second image, a second block of pixel values provided by said sub-matrix (B0) of the matrix of sensors;
- storing said second bloc of pixel values as a second block of pixel values of the super-resolution image, said second block of pixel values being placed right next to said first block of pixel values in the super-resolution image along the image axis (X, Y) corresponding to said one displacement axis (x, y).
2. An imaging method for obtaining an super-resolution image (22) of an object (5) according
to claim 1, wherein the sub-matrix (B0) comprises only one sensor.
3. An imaging method for obtaining an super-resolution image (22) of an object (5) according
to claim 1 or 2, wherein a sensor of the matrix is selected as sensor of the sub-matrix
(B0) in a preliminary step only if said sensor captures a part of the image of the focused
illumination beam (14) on the support plate (6).
4. An imaging method for obtaining an super-resolution image (22) of an object (5) according
to claim 3, wherein the sensor(s) of the matrix is/are selected as sensor(s) of the
sub-matrix (B0) in a preliminary step as the one(s) among the sensors of the matrix capturing a
part of the image of the focused illumination beam (14) on the support plate wherein
the image is the most in focus.
5. An imaging method for obtaining an super-resolution image (22) of an object (5) according
to any one of claims 1 to 4, comprising a step of:
- applying a background removal to the pixel values of the super-resolution image and/or including a step of cropping said pixel values that are outside at least one specified range of upper intensity values, in order to remove from the pixel values related to an individual fluorophore the overlapping intensities of neighbouring fluorophores.
6. An imaging method for obtaining an super-resolution image (22) of an object (5) according
to any one of claims 1 to 5, wherein the object includes at least one fluorescence
source, and the illumination source (1) is adapted to produce an illumination beam
for generating fluorescence from the illuminated object, the image of the target region
captured by the digital camera (9) being an image of the fluorescence generated by
the illuminated object, the obtained super-resolution image being a super-resolution
fluorescence image.
7. An imaging system for obtaining an super-resolution image (22) of an object (5), said
imaging system including a controller (10) and an optical microscope (21), said optical
microscope (21) including :
- a support plate (6) for bearing an object to be imaged,
- an illumination source (1) adapted to produce an illumination beam for illuminating the object,
- an optical element (4) for focusing the illumination beam on the support plate, the section of the object on the support plane currently illuminated by the focused illumination beam (14) being called hereafter the target region,
- a digital camera (9) including a matrix of sensors for capturing an image of the target region, each sensor providing for a respective pixel value, and
- a displacement block (11) for displacing the support plate relative to the focused
illumination beam and to the digital camera along at least two displacement axes among
three axes x, y, z perpendicular to each other, wherein the axes x and y define the
plane of the support plate, said at least two displacement axes (x, y) defining two
corresponding perpendicular image axes of the super-resolution image (22);
said imaging system being characterized in that the controller (10) is adapted for iterating the following operation's set of commanding
the digital camera to capture a first image of the target region currently illuminated
by the focused illumination beam (14), extracting, from the captured first image,
a first block of pixel values provided by a sub-matrix (B0) of the matrix of sensors, storing said first bloc of pixel values as a first block
of pixel values of a super-resolution image of the object, then commanding the displacement
block to displace, by a sub-diffraction limited distance, the support plate along
one of the displacement axis, then commanding the digital camera to capture a second
image of the target region currently illuminated by the focused illumination beam
(14), extracting, from the captured second image, a second block of pixel values provided
by said sub-matrix (B0) of the matrix of sensors, and storing said second bloc of pixel values as a second
block of pixel values of the super-resolution image, said second block of pixel values
being placed right next to said first block of pixel values in the super-resolution
image along the image axis (X, Y) corresponding to said one displacement axis (x,
y).
8. An imaging system according to claim 7, wherein the sub-matrix (B0) comprises only one sensor.
9. An imaging system according to claim 7 or 8, wherein the controller is adapted for
selecting a sensor of the matrix as sensor of the sub-matrix (B0) only if said sensor captures a part of the image of the focused illumination beam
(14) on the support plate (6).
10. An imaging system according to claim 9, wherein the controller is adapted for selecting
sensor(s) of the matrix as sensor(s) of the sub-matrix (B0) in a preliminary step as the one(s) among the sensors of the matrix capturing a
part of the image of the focused illumination beam (14) on the support plate wherein
the image is the most in focus.
11. An imaging system according to any one of claims 7 to 10, wherein the controller is
adapted for applying a background removal to the pixel values of the super-resolution
image and/or for cropping said pixel values that are outside at least one specified
range of upper intensity values, in order to remove from the pixel values related
to an individual fluorophore the overlapping intensities of neighbouring fluorophores.
12. An imaging system according to any one of claims 7 to 11, adapted for imaging an object
including at least one fluorescence source and for obtaining a super-resolution fluorescence
image, wherein the illumination source (1) is adapted to produce an illumination beam
for generating fluorescence from the illuminated object, the image of the target region
captured by the digital camera (9) being an image of the fluorescence generated by
the illuminated object.
REFERENCES CITED IN THE DESCRIPTION
This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.
Patent documents cited in the description
Non-patent literature cited in the description
- KLAR et al.Fluorescence microscopy with diffraction resolution barrier broken by stimulated emissionProc Natl Acad Sci USA., 2000, vol. 97, 158206-10 [0007]
- WANG et al.Science, 2011, vol. 333, 1445-1449 [0134]
- DONNERT et al.Proc Natl Acad Sci USA., 2006, vol. 103, 3111440-5 [0135]